* ♻️ Refactor apply token test to match new render
* ♻️ Refactor crud token test with new render
* ♻️ Refactor general token tes tto use new render
* ♻️ Refactor remapping token tests to use new render
* ♻️ Refactor token set tests to use new render
* ♻️ Refactor token theme tests to use new render
* ♻️ Refactor token tree tests to use new render
The font specific error string was never added to en.po (my own mistake).
Looking further into it, there is no need to add more work to
translators when a generic error goes a long way.
Specially since this is not expected to happen.
* 📎 Fix spelling errors
* 🚧 Temporary workaround for sizing options not working
Add instructions explaining that FlexLayout sizing options do not work.
Relates to https://github.com/penpot/penpot-mcp/issues/39
* 🚧 Temporary workaround for Token resolvedValue not working
Instruct LLM to not use this property.
To be reverted once #8341 is fixed.
* ✨ Improve description of token values
* ✨ Make clear that ExecuteCodeTool serialises automatically
LLMs sometimes decide to apply serialisation themselves, which is unnecessary,
and which this seeks to prevent.
* 🚧 Temporary workaround for fills/strokes being read-only
Add instructions to make the limintations.
Once #8357 is resolved, this can be reverted.
* ♻️ Move high-level instructions to the end
In this way, they can reasonably reference the more low-level concepts
* 📚 Add instructions on cloning and the branch to use
* 📚 Revise instructions on prerequisites
* Do not state that pnpm must be available after Node.js installation
(it is installed by corepack)
* Do not state that caddy is required; it is required only when
rebuilding the API documentation for the server, which is not
a task relevant to regular users.
* Do not strongly suggest that MCP users should be using the devenv.
* Windows: Add pointer to use Git Bash
* 📚 Remove unnecessary details on what the boostrap script does
* 📚 Update information on repository structure
* 📚 Add section on 'Development' to README
